Research About The Progressive Transmission Of 3D Model Based On Regions Of User’s Attention

With the development of network technology and computer graphics,3D model has wide application in many fields, such as virtual reality, medical stereo images and 3D film industry and so on. Recently, models are highly complex and detailed with the development of 3D scanning device and relevant technology. The complexity of 3D model data results in tremendous pressure of network transmission. In other words, the speed of current internet is far from the real-time transmitting requirement of 3D model. An effective method to solve this problem is to use the progressive transmission of 3D model.The progressive transmission of 3D model decomposes the model by multi-resolution to get a basic model and a series of details, and then sends the basic model and the details. Client can receive the basic model quickly to display and operate without waiting for a long time. With the continue transmission of the details, the quality of the received model improved gradually..In the progressive transmission process of 3D model, the basic model is produced by the simplification of 3D model. Important visual features in current mesh simplification was easily lost, which thus often caused distortion of vision. In order to get the basic model with high fidelity, this paper proposes an improved edge-collapse mesh simplification algorithm by introducing Gauss curvature to QEM algorithm. The presented algorithm keeps the important characteristics and improves the fidelity of the basic model.In addition, current simplification algorithms only focus on the fidelity of the whole model, ignoring the significance of the integrity of the sharp regions. According to the study of psychophysics, in the recognition of an object, people concern the overall outline of the object first and then depend on the segmentation, in other word; people will split a complex object into several simple parts. However, current simplification algorithms do not keep the local characteristics of the model effectively, so the basic model simplified does not meet the human cognitive process. Thus, based on the edge-collapse mesh simplification method based on Gauss curvature, the paper proposes a novel 3D model simplification algorithm with lossless sharp regions. The presented algorithm segments the model into several sub-blocks and simplifies each sub-block using the edge-collapse mesh simplification method based on Gauss curvature independently. Experimental results show that the proposed algorithm keeps the sharp regions lossless while ensuring the high fidelity of the whole model.At last, it is necessary to restore the simplified information to get details for restoring the model. However, current methods ignore the importance of restoring the regions of user’s attention primarily, causing low transmission efficiency. Based on the 3D model simplification algorithm with lossless sharp regions, this paper proposes a novel progressive transmission algorithm based on the regions of user’s attention. The presented algorithm predicts the user’s attention degree of each sub-block and restores the simplification information in order depending on the degree.